Vol. 34, No. 6, June 1969
TETRAHYDRO-l14A-DIMETHYL-2-NAPHTHALENONE 1949
Conformational Studies. 111. Consequences of the Conjugate Addition of Cyanide Ion to Rigid Bicyclic Systems. B. Tetrahydro-1,4a-dimethyl-Z-naphthalenone1 OSCAR R. RODIGAND NORMAN J. JOHNSTON Department of Chemistry, University of Virginia, Charlottesville, Virginia 22905 Received July 24, 1968
The dienone named in the title underwent both 1,6 and 1,4 addition when treated with potassium cyanide in aqueous ethanol. In the absence of ammonium chloride, the reaction yielded two lactamols; while in the presence of this salt the major products were three isomeric ketodinitriles. The structures of these substances were determined from their dipole moments and nuclear magnetic resonance spectra, used in conjunction with classical chemical correlations.
I n studies conceived to investigate an improved route for the total synthesis of sesquiterpenes having eudesmane ring structure 1, the feasibility of adding potassium cyanide in a 1,6 fashion across the tetrahydro-l14adimethyl-2-naphthalenone system (2) was examined. The product from such an addition would be ketonitrile 3, an interesting intermediate for the conduct of further transformations. Although 1,6 additions with ketone 2 have been reported previously, the yields have often been poor.2 Furthermore, the nucleophiles were of the malonate type whereas our approach required the attachment of an intrinsically positive carbon to position 7. Thus the sterically small cyanide ion appeared to be an ideal nucleophile for our purposes.
I
2
3
The dienone and enone systems represented in 2 and 3, respectively, have highly characteristic ultraviolet absorption spectra which provided a convenient method for monitoring the addition reactions. The 1,6 addition of potassium cyanide to ketone 2 was attempted in media both with and without the addition of ammonium chloride. I n each case, the disappearance of the dienone chromophore signaled the occurrence of a reaction, but the expected enone absorption never appeared, indicating that ketonitrile 3 was not present in appreciable amount at any one time. Furthermore, the reactions were accompanied by considerable decomposition, but the products could nevertheless be easily isolated by careful chromatography. (1) (a) Taken in part from the dissertation of Norman J. Johnston submitted for the Doctor of Philosophy degree, University of Virginia, 1963;
(b) for part A of this series, see 0. R. Rodig and N. J. Johnston, J . Ore. Chem., S4, 1942 (1969). (2) Y. Abe, T. Harukawa, H. Ishikawa. T. Miki, M. Sumi, and T. Toga, J . Amer. Chem. Soc., 78, 1416, 1422 (1956); P. Dutta, J. K. Chakrabarti, and P. C. Dutta, Chem. Ind. (London), 170 (1955); F. J. McQuillin, ibid., 311 (1954); J. K. Chakrabarti, P. Dutt, and P. C. Dutta, J . Chem. Soc., 4978 (1956); F. J. McQuillin, ibid., 528 (1955); F. D. Gunstone and A. P. Tulloch, ibid., 1130 (1955); J. R. Mahajan, P. Dutt, and P. C. Dutta, ibid., 5069 (1957): T. Miki, J . Pharm. SOC.Jap., 7 5 , 395 (1955); T. Harukawa, ibid., 75, 521 (1955); M. Matsui, K. Toki, S. Kiyamara, Y.Suzuki, and M. Hamuro, Bull. Chem. Soc. Jap., 27, 7 (1954); M. Yanagita, S. Inayama, M. Hirakura, and F. Seki, J . Ore. Chem., 28, 690 (1958).
In the absence of ammonium chloride, the reaction yielded lactomols 4 (42y0) and 5 (6y0) , the structures (Scheme I) of which were supported by their infrared spectra (see Experimental Section) .3 Additional information on the structures of 4 and 5 was forthcoming from acetylation studies. By a method described previously,lb using an acetic anhydride-p-toluenesulfonic acid mixture, cyanolactamol 4 could be either mono- or diacetylated to give '7 and 8, respectively. Amidolactamol5 yielded monoacetyl derivative 9 when treated with this reagent at 35". Under reflux temperatures however (diacetylation conditions) , dehydration of the amide group occurred as well, giving diacetyl derivative 8. Two methods were used to hydrolyse the C-7 nitrile group of lactamol 4 to an amide function. The first employed boron trifluoride in acetic acid4 and yielded the same acetylated amidolactamol 9 obtained from 5. The second method used 90% sulfuric acid and gave a new amidolactamol 6.5 To aid in the further elucidation of the above structures, it was desirable to obtain the corresponding dinitrile precursors. Dienone 2 was therefore treated with potassium cyanide in the presence of ammonium chloride.6 Five substances were isolated : three ketodinitriles (10, 11 and 12) in 3, 8 and 18% yields, respectively, 1% ketonitrile 13, a trace of a ketonitrile of unknown structure, and 29% (by weight) uncrystallizable oil.' Unfortunately, ketonitrile 13 was obtained in sufficient quantity t o allow its complete structural characterization.8 Each of the ketonitriles was hydrolyzed in 3y0 aqueous potassium carbonate. trans compounds 10 and 11 both gave the previously obtained lactamol 4 while cis-ketodinitrile 12 yielded a new cyanolactamol 17. I n addition, compound 11 was converted into ketal 16 which was then treated with potassium carbonate. Acid hydrolysis of the reaction product yielded ketodinitrile 10. No epimerization a t C-7 occurred during (3) The spectrum of 4, taken in dilute chloroform solution, exhibits what appears to be a concentration-dependent band in the amide I1 region, which may indicate a partial opening of the lactamol system. Lactamol 5 was found to be too insoluble to permit such a study. (4) C. R. Hauser and D. S. Hoffenberg, J. Ore. Cfiem., 20, 1448 (1955). (5) Attempts to convert the amide group in lactamol 6 to a nitrile group using either phosphorus pentoxide or thionyl chloride were unsuccessful. (6) W. Nagata, Tetrahedron, 18, 278 (1961). (7) Yields given have been corrected for 51% of recovered ketone 1. (8) This compound was isolated from the mother liquors obtained from the recrystallization of ketodinitrile 10 which contains an equatorial '2-7 nitrile group. I t is possible that 18 may have been formed from 10 by t h e a elimination of hydrogen cyanide, in which case the C-7 nitrile group should retain its equatorial configuration. Attempts to prepare 18 by heating 10 with silver nitrate in aqueous ethanol were unsuccessful.
The Journal of Organic Chemistry
1950 RODIGAND JOHNSTON SCHEME I
6
9 TsOH 350
5
EtOH,
+m 0a : CN + 10
11
*mm : CN
+ - 0
12
13
* p sI%cr 16
& oH
C”
CN.
CN
B,9.05 D
A,hOlD
17
hydrolyzed in relation t o the one a t C-7 demonstrates the important role the carbonyl group plays in this r e a c t i ~ n . ~This is further illustrated in the epimer4zation study involving structure 16 which lacks the carbonyl group. The conditions used for this isomerization were the same as those which effected hydrolysis in the ketodinitriles. Stereochemistry of Ketodinitriles 10, 11 and 12.The elucidation of the stereochemistry of the products obtained from the cyanide addition reactions discussed above pivoted on the structure proofs of the ketodinitriles. The stereochemistry of each ketodinitrile was determined by nuclear magnetic resonance and dipole moment measurements. The structures possible for these compounds (neglecting for the moment the C-1 methyl group and boat or twist conformations of the cyclohexane rings) are shown in Figure 1. The two isomers possible for the rigid trans-fused ring system are A and B, but for their counterparts in the flexible cis-fused systems (C and D) it is necessary to consider the alternate conformational forms C’ and D’ as well.lo The shielding effects of the carbonyl and C-8a nitrile groups on the C-1 hydrogen and the C-1 and C-4a methyl groups had been previously observed on ketonitriles 14 and 15 and a correlation was established between the chemical shifts of these groups and their
zH3cf9
N
0
C, 5.01 D
0
C: 6.28 D
Nc!jfy
0
D,7.50 D Di 5.01 D Figure 1
the ketalization Drocess as was shown bv the acid hydrolysis of 16 gack to ketodinitrile 11. The erne with which the angular nitrile group is
(9) Cf.ref Ib and references cited therein. (10) Cf. E. L. Eliel. N. L. Allinger, S. J. Angyal, and G . A. Morrison. “Conformational Analysis,” Interscience Publishers, New York, N. Y . , 1965,
,,231ff.
Vol. 3 4 , X o . 6, June 1969
TETRAHYDRO-~,~A-DIMETHYL-~-NAPHTHALENONE
NMRDATAFOR Compound
THE
Ring juncture
1951
TABLEI KETODINITRILES A N D KETONITRILES~ C-4a methyl, pprn (angle, d e d c
C-1 methyl? pprn (angle, deg)'
C-1 hydrogen? ppm (angle, deg)c
1.20 (60) 2 . 5 1 (180) trans 1.30 (180) Ketodinitrile 10 1.18 (60) 2.49 (180) trans 1.23 (180) Ketodinitrile 11 cis 1.50 (60) 1.18 (60) 3.08 (60) Ketodinitrile 12 trans 1.24 (180) 1.16 (60) 2.50 (180) Ketonitrile 14d cis 1.48 (60) 1.16 (60) 3.05 (60) Ketonitrile lSd The spectra were determined a t 60 Me in deuteriochloroform solutions using tetramethylsilane as an internal reference. The The geometrical center positions of the C-1 chemical shifts are measured from TMS (0 ppm) in the direction of decreasing field. The angle reported in each parentheses is t,he approximate dihedral methyl doublet and the C-1 hydrogen quartet are reported. See ref lb. angle which that respective atom or group makes with the C-8a nitrile group.
orientations with respect to the keto and nitrile groups (Table I) .lb -4 similar analysis of the C-4a methyl positions in the three ketodinitriles 10, 11 and 12 thus makes it possible to determine whether they have a cisor trans-fused ring system. The data in Table I show that the C-4a methyl peaks of ketodinitriles 11 and 12 agree closely with those observed in trans- and cis-ketonitriles 14 and 15,
14
15
respectively, and thus allow the ring-juncture assignments to be made as shown. The agreement is less precise for ket'onitrile 10, but its value nevertheless corresponds considerably more closely with that of the trans rather than the cis structure (cf. also the C-1 proton values). That 10 and 11 do indeed have the same ring juncture stereochemistry has already been shown by the correlabion of their structures through the ketal 16, and by t,he conversion of both substances to lactamol 4. The configurations of the C-7 nitrile groups in compounds 10, 11 and 12 were determined by dipole moment measurements. The calculated dipole moments expressed in Debye units for the structures shown in Figure 1 are given below the formulas." The experimentally determined dipole moments for the two trans compounds 10 and 11 were 4.96 f 0.05 and 8.06 f 0.08 D, respectively. Thus structure A is confirmed for ketonitrile 10. The low dipole moment value obtained for ketodinitrile 11 remains to be explained. If the ring containing the two nitrile groups exists in the traditional chair conformation, the nitrile groups are oriented 1,3 diaxially. Should sufficient dipole-dipole repulsion exist between these two groups, it may cause appreciable ring distortion or may even force the ring into a boat form. l2 ( 1 1) The dipole nioinents were calculated using the Cartesian coordinate values for the cyclohexylidene ring system as determined by E. J. Corey and R. A. Sneen [ J . Amer. Chem. Soc., 77, 2605 (1955)l and the carbonyl and nitrile vector moments reported by J. M. Lehn, J. Levisalles, and G. Ourisson [Bull. Soc. Chim. Fr., 109F (1963)l (see Experimental Section). (12) Dreiding models show that if this ring is in a twist conformation, i t involves some angle strain. It has also been calculated that in decalins having one ring a boat or twist form, the former should be more stable than the latter IJ. Levisallri. and J. C. N. Ma, Bull. Soc. Chim. Fr., 1597 (1962)l.
The two boat forms possible, lla and llb, should have similar free energies. Unfortunately, insufficient information is available on the energy parameters required for these systems to allow an accurate assessment of their relative stabilities. However, an a priori estimate might favor llb over lla, because the latter has the more severe bow-stern interaction (CH3:H in lla vs. CN: H in llb). In support of this is a calculated13 dipole moment of 8.10 D for conformation llb, in excellent agreement with the observed value. Nevertheless, definitive conclusions cannot be drawn as to the conformation of ketodinitrile 11 because of existing uncertainties as to the effects of mutual interactions of parallel dipoles in close proximity with each other.14 For ketodinitriles 10 and 11 the stereochemistry a t C-1 remains to be assigned. This was made by comparing the chemical shifts of the C-1 hydrogen and methyl peaks with those of the corresponding transketonitrile 14 where the C-1 methyl group has been shown to be equatorial.1b The values (Table I) clearly indicate that this group is also equatorial in ketonitriles 10 and 11, allowing one to make the structural assignments shown in Scheme I.
lla
Ilb
CN
10
B
12
For cis-ketodinitrile 12, the experimental dipole moment of 4.98 f 0.05 D effectively eliminates structures C' and D from c~nsideration.'~From the nmr spectrum of this compound (Table I), the peaks for (13) This dipole moment was calculated using the atom positions as measure from a Dreiding model and is therefore probably less accurate than those obtained using the coordinate system method. (14) Cf. C. Djerassi. R. A. Schneider, H. Vorbrueggen, and N. L. Allinger, J . O w . Chem., 18, 1632 (1963). (15) The ring containing the two 1,3-diaxial nitrile groups in structure C' could prefer a boat conformation so as to relieve the dipole-dipole repulsion between these groups. However, the most stable structure should actually be C where both rings are in chair forms and the C-7 nitrile group is equatorial.
1952
The Jownal of Organic Chemistry
RODIGAND JOHNSTON
the C-1 methyl group (1.18 ppm) and the C-1 hydrogen (3.08 ppm) indicate that the dihedral angle each of these makes with the angular nitrile group is approximately 60". Molecular models readily shown that only structure D' (C-1 methyl equatorial or axial) satisfies these arrangements. Furthermore, only when the C-1 methyl group is assigned an equatorial position in D' is its relationship to the carbonyl group the same as that in the other compounds cited in Table I. Hence, ketodinitrile 12 can be assigned the structure shown above. Stereochemistry of Lactamols 4 and 17.-The hydrolysis of trans compounds 10 and 11 established the trans ring juncture of cyanolactamol4. Since the C-7 nitrile group in 4 would be expected to retain the more stable equatorial position, hydrolysis must have involved the epimerization of the axial C-7 nitrile group in 11.16 Likewise, the C-1 methyl group which is equatorial in both 10 and 11 would be expected to remain so in product 4. This assignment is in accord with the nmr spectrum of the acetyl derivative 7 which exhibits the C-1 methyl and C-1 hydrogen multiplets centered a t 0.91 and 2.41 ppm, re~pective1y.l~ Cyanolactamol 17 obtained from the hydrolysis of ketodinitrile 12 is assumed to have an axial C-1 methyl group by analogy with the corresponding lactamol obtained in the enone series.Ib The configuration of the C-7 nitrile function in 17 remains uncertain. Molecular models show that if this group retains its original configuration, it will be axial and extremely close to the axial met,hyl group a t C-1. I n view of the slightly larger steric requirement for a nitrile group as compared to hydrogen18and the fact that the hydrolysis conditions are sufficient to cause expimerization at C-7, one is encouraged to favor an equatorial conformation for this group in this case. Stereochemistry of Derivatives of Lactamols 4 and 5.-The stereochemistry a t C-1, C-2, C-4a and C-Sa is the same in both 4 and 5 as was shown by their conversion to the same acetate 9. The hydrolysis product 6 obtained from cyanolactamol 4 should also have the same stereochemistry at these centers. Therefore, compounds 5 and 6 must be epimeric at C-7. I n the hydrolysis using 90y0 sulfuric acid, the relatively bulky amido group in 6 would be expected to retain its equatorial configuration. On the other hand, the conditions employed during the acetylation of amidolactamol5 can be expected to effect epimerization of an amido group axial at C-7.19 Hence, this group in lactomol 5 is assigned an axial configurationPo Mechanistic Notes.-The products obtained from the addition of potassium cyanide to conjugated dienone 2 show that 1,6 addition must occur first, followed by a rapid 1,4 addition. The observed disappearance of the dienone chromophore in the ultraviolet spectrum of the reaction mixture as addition proceeded without concomitant increase in the absorption of the enone chromo(16) As indicated previously, such an epimerization occurs under these conditions in the conversion of ketodinitrile 11 to 10 via the ketal 16. (17) Compare with the values observed for these groups in the monoacetates of lactamols 4 and 5 in ref lb. (18) N. L. Allinger and W. Sykrybalo, J . Ore. Chem., 27, 4601 (1962); B. Rickborn and F. R. Jensen, ibid.,4606 (1962). (19) B. Rickborn and F. R. Jensen, J. 0.r~.Chem. 37, 4608 (1962). and references cited therein.
phore supports such a mechanism.21 A rough measure of the rates of the potassium cyanide additions t o both dienone 2 and enone 18 was obtained by quantitatively
18
sampling the reaction mixtures at periodic intervals and observing the respective decreases of the absorption peaks with time (247 mp for the enone and 287 mfi for the dienone). The results indicate that under the conditions investigated, the rates of addition to the enone 18 were about the same in both the presence and absence of ammonium chloride, being approximately 0.4 1. mol-' hr-1. Furthermore, the rates of 1,6 addition to dienone 2 in the presence and absence of ammonium chloride were 10 and 2.5 times slower, respectively, than were the 1,4 additions to enone 18. Consequently, the 1,6 addition step appears to be the ratedetermining one. The reversibility of the steps in this sequence precludes any discussion of the stereochemistry of these conjugate additions. Experimental Section22 Potassium Cyanide Addition to 4,4a,5,6-Tetrahydro-l,4adimethyl-2- (3H) -naphthalenone (2). A. In Aqueous Ethanol.-A solution of 4.0 g (0.023 mol) of dienone 2 (n% 1.5660),w 8.0 g (0.123 mol) of potassium cyanide, 160 ml of 95% ethanol and 16 ml of water was heated a t reflux with stirring for 12 hr. The solvent was removed in vacuo and the remaining orange oil was diluted with water, extracted with ethyl acetate, saturated with sodium chloride and extracted continuously for 24 hr with ethyl acetate. The combined extracts were dried, the solvent was removed in vacuo and the 4.5 g of orange semisolid remaining was chromatographed on Fl~risil.*~ Benzene and benzene-ether ( 1 O : l ) eluents yielded 0.42 g (4%) of unreacted dienone 2 while benzene-ether ( l : l ) ,ether and ether-ethyl acetate (5: 1 ) eluted 2.37 g (42%) of cyanocd(20) The stereochemistry given for the C-7 substituents in all compounds except 10, 11, 12 and 16 is that which is considered to be the most likely from the available evidence and partly rests on the assumption that the substituent is the more stable in an equatorial position. I t is conceivable, however, that an axial C-7 amido group might be stabilized through intramolecular bonding of the type shown in i. A similar argument can be applied to lactamol 4. Infrared dilution studies carried out on amidolactamols 5 and 6 in dioxane were inconclusive with respect to this point.
HII&
HN--c-o--
p"
N& i
(21) Patents issued to A. Bowers and H. J. Ringold describe the 1,0 addition of cyanide ion to a variety of 3-keto A486 steroids. I t is reported that potassium cyanide in absolute or 95% ethanol a t reflux for 2.5 hr afforded the 7n- and 78-cyano epimers. No other reaction products or yields are described [U.S. Patents 3,050,534 (August 21. 1963) and 3,099,664 (July 30, 1903). (22) All melting points were determined in a heated oil bath and are corrected, while boiling points are uncorrected. The nmr spectra were determined in deuteriochloroform solution (unless specified otherwise) on a Varian A-60 spectrometer and chemical shift values are given in parts per million (ppm) measured downfield from tetramethylsilane used as an internal standard. The infrared spectra were determined in the solid state in a potassium bromide matrix (unless indicated otherwise) on a Perkin-Elmer Model 21 spectrophotometer. The ultraviolet absorption spectra were obtained with a Perkin-Elmer Model M O A Spectracord. The microanalyses were performed by Mrs. D. Ellis and Mrs. W. Coyne of this laboratory. (23) D. K. Banerjee and V. B. Angadi, J . Ow. Chem., 26, 2988 (1961). (24) Floridin Company, Tallahassee, Fla., 60-100 mesh.
Vol. 34, N o . 6 , June 1969
TETRAHYDRO-~,~A-DIMETHYL-~-NAPHTHALENONE 1953
tam014 which was recrystallized from ethyl acetate to give 1.46 g (26%) of colorless crystals, mp 210-214". Further recrystalliaation from ethyl acetate raised the melting point to 215-216"; ir 3420 (OH), 3176 (NH), 3058 (sh, b, lactam), 2232 (C-N) and 1681 cm-1 (lactam ( 3 4 ) ; ir (chloroform) 3534 (OH), 3390 (NH), 2232 ( C s N ) , 1698 (lactam C 4 ) and 1597 cm-l (-CONH2 amide 11); no selective ultraviolet absorption was observed above 220 mp with a 1% ethanol solution. Anal. Calcd for CI~HZONZOZ: C, 67.71; H, 8.12; N, 11.28; mol wt, 248. Found: C, 67.65; H, 8.25; N, 11.47; mol wt, 248 (Rast). Cyanolactamol 4 was insoluble in hot water, hot 5% sodium bicarbonate solution, and hot 10% hydrochloric acid solution but was soluble in cold 10% sodium hydroxide solution, and failed to react with dinitrophenylhydrazine reagent."* Further elution with ethyl acetate gave 0.08 g (2% by weight) of a water-soluble solid of unknown structure, mp 220-248" dec, which crystallized from ethyl acetate as a colorless solid: mp 248-256' dec; ir 3441, 3349, 3199, 3090 (sh), 2247, 1723 (sh, b) and 1678 cm-1 (b) ; unsaturation tests were negative.,$b Anal. Found: C, 60.87; H, 7.19; N, 10.42; mol wt, 297 (Rast). Finally, elution with acetone yielded a water-soluble brown solid which was washed with ethyl acetate, yielding 0.20 g (3%,) of amidolactamol 5 as colorless crystals: mp 250-252" dec; ir 3391 (OH, NH2),3206 (NH, NH,), 3102 (lactam), 1690 (lactam C 4 ) , 1657 (b, C=O amide I ) and 1599 cm-1 (NH, amide 11); no selective ultraviolet absorption was observed above 220 mp with a 1% ethanol solution; n m P (DMSO-&) 6 0.78 (d, 3, J = 7 Hz. C-1 CH3). 0.97 (s, 3, C-4a CHI), 6.50 (s, 1, NH), 7.27
needles: mp 152-152.5"; uv max 287.5 m r ( e 19); ir 2242 (CmN) and 1721 cm-l ( C = O ) ; nmra .3 1.18 (d, 3, J = 7 Hz, C-1 CHa), 1.50 (8, 3, C-4a CHa), 3.08 (qr, 1, J = 7 Hz, C-1 H). Anal. Calcd for CI'HlsN20: C, 73.01; H, 7.88; N, 12.17. Found: C, 72.98; H, 7.94; N, 12.14. Continued elution with benzene-ether ( 5 : l ) and ether produced a white solid which was recrystallized from benzene, giving 0.21 g (3%)**of ketodinitrile 10, mp 214-226". Further recrystallization from benzene raised the melting point to 225-228'; ir 2237 (C-N) and 1706 cm-l ( C 4 ) ; n m F 8 1.20 (d, 3, J = 7 Hz, C-1 C&), 1.30 (s, 3, C-4a CHa), 2.51 (qr, 1, J = 7 He, C-1 H). Anal. Calcd for C14H18N20:C, 73.01; H, 7.88; N, 12.17. Found: C, 72.90; H, 8.13; N, 11.99. The mother liquor yielded 68 mg (l%)z8 of cyanoenone 13, mp 90-150". Recrystallization of this material from carbon tetrachloride gave colorless crystals: mp 97-98', uv max 244.5 mp (log c = 4.119) ; ir 2237 (C=N), 1664 (C=O) and 1612 cm-1 (C=C). Anal. Calcd for CiaHnNO: C, 76.81; H, 8.43; N, 6.89. Found: C, 76.87; H, 8.29; N, 7.06. Elution with ether and ether-ethyl acetate (7: 1) yielded 0.18 g of colorless crystals which, after recrystallization from ethyl acetate, gave 30 mg of a ketonitrile of unknown structure: mp 203-204"; ir 3483, 3344, 2235 and 1700 cm-l. Anal. Calcd for Cl~H1,NOa: C, 65.80; H, 8.07; N, 5.90. Found: C, 65.82; H, 8.24; N, 6.24. It formed a yellow dinitrophenylhydrazone the infrared and qualitative uv spectra of which were similar to those expected of a dinitrophenylhydrazone of a saturated ketone. After ether-ethyl acetate (7: 1, 1 : 1) had removed 0.44 g (4% by weight) of dark red oil which resisted all attempts a t crystallization, ether-ethyl acetate (1 : 1) and acetone eluted 0.76 g of brown solid which was recrystallized several times from ethanol, yielding 0.48 g (8%)2* of ketodinitrile 11 as colorless plates: mp 182.5-183.5"; ir 2237 (C=N) and 1704 cm-1 ( C 4 ) ; nmr26 6 1.18 (d, 3, J = 7 Hz, C-1 CH3), 1.23 (s, 3, C-4a CHa), 2.49 (qr, 1, J = 7 Hz, C-1 H). Anal. Calcd for C I ~ H ~ ~ NC,~ O73.01; : H, 7.88; N, 12.17. Found: C, 72.83; H, 7.90; N, 12.20. Further elution with acetone gave 1.00 g (loyoby weight) of dark red oil which failed to crystallize. Hydrolysis of Ketodinitrile 12.-Ketodinitrile 12 (250 mg) was d d e d to a boiling solution of 1.0 g of potassium carbonate in 20 ml of 9570 ethanol and 14 ml of water. The mixture was heated at reflux for 2 hr, evaporated in vacuo and extracted with ethyl acetate. When the combined extracts were dried and evaporated in vacuo, 207 mg (77%) of solid remained which was triturated with ether yielding 178 mg (66%) of cyanolactamol 17, mp 208-230". Recrystallization of this material from ethanol raised the melting point to 254-257" dec; ir 3333 (OH), 3190 (b, NH), 2242 (CEN) and 1664 cm-l (b, lactamol C=O). Anal. Calcd for CldH10N202: C, 67.71; H, 8.12; N, 11.28. Found: C, 67.41; H, 8.39; N, 11.25. Hydrolysis of Ketodinitrile 10.-Using the procedure given above for ketodinitrile 12, 34 mg of 10 produced 33 mg of solid which was chromatographed on Fl~risil.*~ Elution with benzene gave 4.4 mg (15y0) of impure ketonitrile 13, uv max 244 mp ( e 5900). Ethyl acetate eluents afforded 18 mg (49%) of cyanolactamol 4, mp 208-211", whose infrared spectrum was identical with that of authentic 4. Recrystallization of this material from ethyl acetate raised the melting point to 214-215" and a mixture melting point with authentic 4 was undepressed. Hydrolysis of Ketodinitrile 11.-Following the above procedure, 84 mg of 11 gave 70 mg of oil which was chromatographed on F10risil.*~ Again, two major fractions were obtained. Elution with benzene yielded 17 mg (23%) of ketonitrile 13, uv max 244 mp ( e 8400), the infrared spectrum of which indicated that about 20% starting material 11 was also present. Ether and ethyl acetate eluents gave 36 mg (40%) of cyanolactamol4 whose infrared spectrum was identical with that of authentic 4. One recrystallization of this material from ethyl acetate raised the melting point to 214-215" and a mixture melting point with authentic 4 was undepressed. Treatment of Cyanolactamol4 with Acetic Anhydride. A. At Room Ternperature,*g Cyanolactamol Acetate (7) .-A solution
(8,
2,'"~).
'
Anal. Calcd for C~~HZZNZOI: C, 63.13; H, 8.33; N, 10.52. Found: C. 63.42: H. 8.51: N. 10.56. Recrystailization of amidolactamol 5 from isopropyl alcoholpetroleum ether gave a colorless solid containing alcohol of crystallization. Its melting point in an open capillary was below 200" but on further heating the sample resolidified and melted again a t 251-252.5" dec. Drying in vacuo at 140" for 24 hr removed the solvent of crystallization. A combined 0.66 g (16.5% by weight) of unidentifiable highly colored oils were also obtained at several points in the elution spectrum of the above chromatogram. When the above experiment was scaled up 2.5 times and a 19.5hr reflux period was employed, the following yields were obtained: dienone 2, 8%; cyanolactamol 4, 26%; amidolactamol 5, 4%; a lactamol derived from enone 18:' 0.7%; and unidentifiable highly colored oils, 14% by weight. The reaction rate was monitored by removing aliquot parts of the reaction solution a t periodic intervals and observing the decrease of the 288-mp absorption maximum in the ultraviolet spectrum of dienone 2. B. In Aqueous Ethanol with Ammonium Chloride.-A solution of 10.00 g (0.0567 mol) of dienone 2, n% 1.5660, in 100 ml of 95y0 ethanol was added to a solution of 14.75 g (0.227 mol) of potassium cyanide and 10.60 g (0.198 mol) of ammonium chloride in 100 ml of water. The mixture was heated to the reflux temperature and clarified with an additional 50 ml of ethanolwater ( 3 : 2 v/v). After 6 hr, the concentrations of both salts were doubled and the heating a t reflux was continued for another 8.5 hr. The rate of conjugate addition was followed by periodic ultraviolet sampling as previously described for the untempered reaction. The solvent was removed in vacuo and the brown residue was extracted with ethyl acetate, diluted with water and extracted continuously with ethyl acetate for 24 hr. The combined extracts were dried and the solvent was removed in vacuo leaving 10.8g of dark oil which was chromatographed on Florisil.24 Elution with benzene and benzene-ether (20: 1) yielded 4.04 g of starting material 2. Benzene-ether (20: 1, 5: 1) eluents gave 2.17 g of white semisolid which was separated by crystallization from benzene into 1.04 g of dienone 2 (51% total recovery) and 1.13 g of ketodinitrile 12, mp 149.5-151". A recrystallization of 12 from benzene yielded the substance as colorless (25) R. L. Shriner, R. C. Fuson, and D. Y. Curtin, "The Systematic Identification of Organic Compounds," 5th ed, John Wiley & Sons, Inc., New York, N. Y.,1984: (a) p 126; (b) pp 121, 149. (26) The chemical shift values for all doublets and quartets were measured at the geometrical midpoint between the peaks. (27) This material was formed from a small amount of enone 16 impurity in the starting material ( ~ B D 1.5651), and is lactamol 5 in ref lb. (28) Yields are adjusted for recovered starting material.
(29) Huang-Minlon, E. Wilson, N. L. Wendler, and M. Tishler, J . Amer. Chem. Sac., 74, 5394 (1952).
1954 RODIGAND JOHNSTON of 99 mg of cyanolactamol 4, 90 mg of p-toluenesulfonic acid monohydrate and 5 ml of acetic anhydride was stirred a t 35" for 24 hr, cooled, poured onto crushed ice and allowed to warm to room temperature. The resultant clear solution was extracted with ether, saturated with sodium chloride and further extracted with ether. The combined extracts were washed with saturated sodium bicarbonate solution until evolution of gas had ceased, dried and the solvent removed in vacuo. The residue consisted of 114 mg of cyanolactamol acetate 7 as colorless crystals, mp 180-185'. When this material was recrystallized from benzene 88 mg (76%) was recovered: mp 185.5-186.5'; ir 3333 (N-H), 2237 ( C s N ) , 1730 (CH&=O), 1709 (lactam C = O ) , 1259, 1225 and 1212 cm-l (CH&OO-) ; nmra66 0.91 (d, 3, J = 7 Hz, C-1 CHs), 1.10 (s, 3, C-4aCHs), 2.07 (s, 3, CHsCO), 2.41 (qr, 1, J = 7 Hz, C-1 H ) , 5.15 (s, 1, NH). Anal. Calcd for CI.~&ZNZOI:C, 66.18; H, 7.64; N, 9.65; mol wt, 290. Found: C, 65.89; H, 7.51; N, 9.85; mol wt, 286 (Rast). Hydrolysis was carried out by heating at reflux 109.6 mg of 7 for 12 hr in 25.24 ml of 0.0393 N potassium hydroxide solution. Titration of the cooled mixture with 12.12 ml of 0.0488 N hydrochloric acid solution to a phenolphthalein end point furnished a saponification equivalent of 268 (theoretical 290). To isolate the reaction product, the solution was acidified and extracted with ethyl acetate both before and after concentration in vacuo. The combined extracts were washed with saturated sodium bicarbonate solution, dried and the solvent removed in vacuo. Upon trituration with ethyl acetate-petroleum ether (bp 30-60"), the residual oil gave 34 mg (36% recovery) of cyanolactomol 4 mp 214-215", the infrared spectrum of which was identical with that of an authentic sample of 4. B. At Reflux Temperature. Cyanolactamol Diacetate ( 8 ).A solution of 0.51 g of cyanolactamol4,0.80 g of ptoluenesulfonic acid monohydrate and 25 ml of acetic anhydride was heated a t reflux for 24 hr. The resulting dark solution was processed in the manner described above for the room temperature acetylation, yielding a dark oil which solidified on standing. This material was chromatographed on Florisilz4where elution with benzene and benzene-ether ( 1 : l ) gave 0.48 g (7101,) of the cyanolactamol diacetate 8 in several fractions melting over a range of 149-156'. Recrystallization from ethyl acetate-petroleum ether gave 8 as light yellow plates: mp 156-157"; ir 2242 (C=N), 1739 (sh, CHaCON-), 1727 (b, CH&OO-), 1708 (b, lactam C = O ) and 1274, 1261, 1345 (sh), 1235-1217 (b, doublet) and 1203 cm-1 (sh, c H s c o o - ) . Hvdrolvsis of Cvanolactamol 4. A. With 90% Sulfuric Acid.w b d o l a c t & o l 6.-A solution of 510 mg of 4"and 15 ml of 90% sulfuric acid was heated with stirring for 1 hr a t 125". The yellow reaction mixture was poured onto crushed ice, warmed to room temperature and the resultant solution extracted with ethyl acetate. The aqueous layer was then carefully neutralized with dilute base, extracted with ethyl acetate, evaporated to dryness in vacuo and again extracted with ethyl acetate. The combined extracts were dried and evaporated in vacuo yielding an oil. Trituration of the oil with acetone yielded 100 mg (19%) of amidolactamol 6 as colorless crystals, mp 238-242' dec, with gas evolution, and also gave a filtrate containing 150 mg of oil (further treated as described below). Recrystallization of the former product from isopropyl alcohol gave crystalline 6: mp 244-245" dec; ir 3448 (OH), 3226 (very b, "2, NH, lactam), 1717 (lactam C=O), 1656 (b, amide I C=O) and 1609 cm-l nmrz6 (DMSO-&) 6 0.78 (d, 3, J = 7 Hz, (sh, amide I1 "2); C-1 CH3), 0.97 (s, 3, C-4a CHI), 7.37 (s, 2, "2). Anal. Calcd for Cl4HZ2N203:C, 63.13; H, 8.33; N, 10.52. Found: C, 63.29; H, 8.34; N, 10.61. The dry salt remaining from the above trituration was extracted with acetone giving 0.29 g of oil which was combined with the oil from the above filtrate and triturated with isopropyl alcohol. An additional 184 mg (34%) of 6 was obtained in several crops having melting point ranges from 232-238" dec. B. With Boron Trifluoride-Acetic Acid.4 Amidolactamol Acetate (9).-Cyanolactamol 4 (0.15 g) was added to 2 ml of glacial acetic acid saturated with boron trifluoride. The solution was heated for 10 min a t 120", cooled, carefully neutralized with 10% sodium hydroxide solution, extracted with ethyl acetate, then made alkaline with dilute base and again extracted with (30) W. J. Hickenbottom, "Reactions of Organic Compounds," Longmans, Green and Co., New York, IY., . 1957, pp 382-383.
The Journal of Organic Chemistry ethyl acetate. The combined extracts were dried and the solvent was evaporated in vacua yielding an oil. When this material was triturated with ethyl acetate-petroleum ether, 0.14 g (75%) .of the amidolactamol acetate 9 was obtained as a colorless solid, mp 214-222". Recrystallization of this solid from acetone followed by filtration through FlorisiP in ethyl acetate gave rise to apparently less pure material, mp 200-207", the infrared spectrum of which was poorly resolved. However, when this materid waa recrystallized from acetone-petroleum ether, colorless crystals were obtained: mp 216-219"; ir 3465 (OH), 3367 (b, NH2), 3185 (NH2, NH), 3086 (lactam), 1734-1704 (b, lactam C=O, CHsC=O), 1664 (amide I C=O), 1621 (amide I1 NI-Ii), 1254, 1221, 1213 and 1202 cm-l (CHICOO-). Anal. Calcd for CMHZ,NZO~: C, 6231; H, 7.84; N, 9.09. Found: C, 62.07; H, 7.72; N, 9.25. Acetylations of Amidolactamol 5.-A solution of 0.23 g of 5, 0.17 g of p-toluenesulfonic acid monohydrate and 10 ml of acetic anhydride was stirred at 35" for 24 hr,as then poured onto crushed ice and allowed to reach room temperature. The solution containing suspended crystalline material was extracted with ethyl acetate, saturated with sodium chloride and again extracted with ethyl acetate. The combined extracts were washed with saturated sodium bicarbonate solution, dried, and the solvent evaporated in vacuo, yielding 0.22 g of a yellow solid. This was recrystallized from ethyl acetate giving 0.13 g (46%) of compound 9 (ir spectrum), mp 211-220'. This colorless solid could not be purified by further recrystallization from ethyl acetate so 91 mg of the substance were heated at reflux for 24 hr with 200 mg of p-toluenesulfonic acid monohydrate in 15 ml of acetic anhydride. The reaction mixture was poured onto crushed ice, allowed to warm to room temperature and extracted with ethyl acetate. The extracts were washed with 5% sodium hydroxide solution, dried and the solvent removed in uacuo. There remained 90 mg of yellow solid which was recrystallized several times from ethyl acetate, giving 21 mg (21%) of the diacetate 8, mp 157.5-158". A mixture melting point of this material with 8, mp 156-157", produced by the action of refluxing acetic anhydride on cyanolactamol4, was undepressed, and the infrared spectra were identical. Anal. Calcd for C18H2&N2O4: C, 65.04; H, 7.28; N, 8.43. Found: C, 64.85; H, 7.32; N, 8.36. Preparation of the Ketodinitrile Ketal 16.-A mixture of 68.0 mg of ketodinitrile 11, 500 mg of ethylene glycol, a few small crystals of p-toluenesulfonic acid, and 5 ml of benzene was placed in a flask attached to a D e a n a t a r k water trap filled with benzene and heated at reflux for 5 hr. The mixture was cooled, 5 ml of ethyl acetate was added, and the solution was then washed in a centrifuge tube with saturated sodium bicarbonate solution and with water. The organic layer was dried over anhydrous sodium sulfate and the solvent was removed under vacuum. There remained 81.2 mg (98%) of crude ketal 16, mp 204-209'. Repeated recrystallization from benzene raised the melting point to 220-224" with previous sintering a t 216": ir 2850, 2210 ( C e N ) , 1440, 1100, 1080, 1045, 953, 927, 900 and 685 cm-1. Anal. Calcd for C1BH22N202:C, 67.17; H, 8.86; N, 11.19. Found: C, 66.98; H, 8.68; N, 11.00. A mixture of 6 mg of the ketal 16,3 ml of ether and 3 ml of 3 N hydrochloric acid was stirred magnetically for 15 min. The mixture was transferred to a centrifuge tube and the aqueous layer removed with a dropper. The organic layer was washed thoroughly with water and dried over anhydrous sodium sulfate. Removal of the ether in vacuo left 3 mg of solid, the infrared spectrum of which was identical with that of ketodinitrile 11. Conversion of the Ketodinitrile Ketal 16 into the Ketodinitrile 10.-Potassium carbonate (1 g) was dissolved in a boiling solution of 20 ml ethanol and 14 ml water. A 5-ml sample of this solution was added to 47.6 mg of ketal 16 and the homogeneous mixture was heated at reflux for 2 hr. The solvent was removed in vacuo and the residue was triturated with water and filtered. The colorless solid remaining on the filter was washed several times with water and dried, yielding 32.1 mg of material, the infrared spectrum of which indicated that a small amount of ketal hydrolysis had occurred. Therefore, this product was not further purified but rather was dissolved in 10 ml of ether and hydrolyzed with 7 ml of 3 N hydrochloric acid using the procedure described earlier. A colorless solid was obtained which was chromatographed on Florisil. Pure benzene eluted a small amount of oil which was not further identified, while elution with 100% ether yielded a combined
Vol. 34, No. 6, June 1969
1955
TETRAHYDRO-1 ,4A-DIMETHYL-%NAPHTHALENONE
TABLEI1 THEORETICAL DIPOLEMOMENTCALCULATIONS' Coordinatesb
Factor
2/,
2%
2,
Directed moments
k,
k,2,
k,Y,
k,2,
Carbonyl oxygen -0.732 0 0.258 3.8524 -2.8200 0 0.9939 Axial nitrile at c or e (as in 10a or 12, form D) -0.031 f0.049 1.09 3.4538 -0.1071 f O . 1692 3.7646 Equatorial nitrile a t c, c' or e (as in loa) 0 577 f0.886 -0.300 3.4302 1.9792 f3.0392 -1.0291 Axial nitrile a t d' (as in in 12, form D) -1.04 0 -0.313 3.4712 -3.6100 0 -1.0865 All values except coordinates are in Debye units. * Origin a t carbon atom bearing the substituent. Coordinates refer to those of attached hydrogen atoms and therefore define the bond directions necessary for the calculation of the dipole moments by the described method. (1
weight of 28.1 mg (72y0) of colorless solid which proved to be identical with ketodinitrile 10 as evidenced by their infrared spectra and melting points. Further elution with 100% acetone gave an additional small amount of noncrystalline material, the infrared spectrum of which indicated that it was possibly a mixture of ketodinitriles 10 and 11. Theoretical Dipole Moment Values. A. Method.-Theoretical dipole moments were calculated for all the possible configurations of the two ketonitriles 14 and 15 and the three ketodinitriles 10, 11 and The bond moments (pi) are the magnitudes of th_e individual group dipoles For each group, a vector (pz) can be defined in terms of the local coordinates of Corey and Sneen" (z,, yl, z , ) for substituents on a cycJohexylidene ring. This vector acts parallel to the dipole vector ( e , )and the magnitude of is related to the magnitude of G by
(z).
lGl=ktlEl
(1)
of the resultant dipole moment p = C(2ktZ1)'
(p)
for the molecule is given by 31
+ (2k,y,)2 + (Zk,z,)'l'~'
(3)
B. Calculations.-The bond moments ( pt) of Lehn, Levisalles, and Ourissonll were employed, i.e., carbonyl, 2.99 D, and nitrile, 3.77 D. The Corey and Sneen" coordinates for the axial and equatorial substituents of all pertinent carbon atoms (see loa, atoms a, c, d, e) are listed in Table I1 along with the calculated values of k , and directional moments. The dipole moments ( p ) of the various configurations of nitrile-containing ketones were calculated substituting the proper directional moment values from Table I1 into eq 3. Two examples will illustrate the facile application of this technique for calculating p from any number of separate group moments." ketodinitrile 10
The factor k , is calculated (for each group) from the relationship
The components of G along the coordinate axes (directed moments) are then given by (k,z,,k,y,, k,z,) and the magnitude
TABLE 111 EXPERIMESTAL DIPOLEMOMENTDATA wL2x 10-8 €12 Ketodinitrile 10 1.19 2.2860 1.75 2.2924 2.85 2.3053 Ketodinitrile 11 1.21 2.3097 2.48 2.3474 3.55 2.3816 Ketodinitrile 12 1.07 2.2855 2.19 2.2990 2.75 2.3052 Ketonitrile 14 1.08 2.2896 2.03 2.3053 2.82 2.3180 3.41 2.3283 Ketonitrile 15 0.73 2.2758 1.45 2.2786 3.15 2.2873 4.10 2.2929
AN
4.63 5.03 5.32 0.50 1.50 2.30 0.36 0.80 1.10 4.54 5.37 6.04 6.19 4.82 5.71 6.75 7.90
TABLEI V EXPERIMEKTAL DIPOLEMOMENTCALCULATIONS Conipd
10 11 12 14 15 (31)
LY
11.63 30.73 11.73 16.61 5.07
V
M z
0.0272 0.0503 0.0288 0.0463 0.0598
230.3 230.3 230.3 205.3 205.3
/*Pa
502 1329 507 640 194
M2
4.96 f 0 . 0 5 D 8.06fO.OSD 4.98 f 0 . 0 5 D 5.59 f 0 . 0 6 D 3.08 f 0 . 0 3 D
K. J. Johnston, Ph.D. Dissertation, University of Virginia, Charlottes-
ville, Va., 1963.
CN 10
H
CH3 108
The directional moments of the carbonyl carbon a, the axial nitrile group on carbon c and the equatorial nitrile on carbon c' are employed. Thus
+
Z k , Z , = -2.8200 - 0.1071 1.9792 = -0.9479 D Z k , y , = 0 - 0.1692 - 3.0392 = -3.2084 D Zk,z, = 0.9939 3.7646 - 1.0291 = 3.7294 D p = [ (-0.9479) (-3.2084)' (3.7294)2]L'2 p = 5.010 D
+ +
+
ketodinitrile 12, form D
CH3
I
NF CN
o a R'
12 (form D), R = H ; R' = CHa or R' = H; R = CH3
Zkiri = -2.8200 - 0.1070 - 3.6100 = -6.5371 D Zkiyi = 0 0.1692 0 = 0.1692 D Z k i ~ i= 0.9939 3.7646 - 1.0865 = 3.6720 D p = [(-6.5371)' (0.1692)' (3.6720)2]1'2 p = 7.500 D
+ +
+
+
+
(32) A similar treatment has been reported by N. L. Allinger, M. A. DaRooge, M. A. Miller, and B. Waegell, J . Org. Chem., 18, 780 (1963); however, in their cage ki was not evaluated from a local coordinate system. (33) For additional examples in this series, see ref 31.
The Journal of Organic Chemistry
1956 BOEKELHEIDE, RAMEY,STURM,MIYASAKA, AND HESS Ketodinitrile 11 in Form 1lb.-Corey and Sneen“ do not furnish the Cartesian coordinates of the atoms in a boat conformation of the cyclohexylidene ring, therefore the directional moments kizi, kiyi and kizi of the C-7 nitrile group for this form were determined from measurements on a Dreiding model. An orthogonal coordinate system was constructed with string and the C-7 carbon atom of the model for form l l b was inserted a t the origin such that the coordinate system with respect to the remaining atoms was identical with that employed by Corey and Sneen.11 The angles the C-7 nitrile group made with the x, y and z axes were carefully measured to be 83, 40 and 52”, respectively. The product of the cosine of the angle times the nitrile group moment gave a new set of directional moments for the C-7 nitrile group in this conformation: 0.46, -2.89 and 2.32 D for the z, y and z coordinates, respectively. These data along with the directional moments of the carbonyl group and the axial nitrile group on carbon c allowed a calculation of 8.1 D for the dipole moment of l l b . DMOl diC. Experimental Dipole Moment Values.-The polemeter manufactured by Wissenshaftlich-Technische Werkstatten was used for the dielectric constant measurements (eI2) a t various weight fraction (weight of sample/weight of solution) concentrations (w12). This apparatus was thermostated to 25 f 0.01” and has been described elsewhere.34 Measurements 2.2725. 1.49790, d26 0.87368, were made in benzene, n541025 Refractive indices (ANn) were measured at the same concentrations ( ~ 1 ~in) a 2.5-cm cell on a Rayleigh interferometer operating (34) N. L. Allinger, H. M. Blatter, M. A. DaRooge, and L. A. Freiberg, J . Ore. Chem., 26, 2550 (1961).
at 25“ on a wavelength of 0.546 X
cm. This instrument is described el~ewhere.~6The e12, AN^^ and w l 2 values are presented in Table 111. The calculations were made essentially by the method of Halverstadt and Kumler3’ as modified by Guggenheim,*’ and are fully described by Smith.88 Details for the calculation of v are given by Bauer, Kajans and Lewha6 The values for a, Y , Mz, pP2 and p2 are listed in Table IV.
Registry No.-Cyanide
ion, 57-12-5; 2, 19291-93-1;
4, 19291-94-2; 5, 19291-95-3; 6, 19291-96-4; 7, 19291-97-5; 8, 19291-98-6; 9, 19291-99-7; 10, 19292-00-3; 11, 19292-01-4; 12, 19292-02-5; 13, 19292-03-6; 14, 19292-04-7; 15, 19292-17-2; 16, 19292-05-8; 17, 19292-06-9.
Acknowledgment.-The authors are most grateful to Jean-Marie Lehn, Institut de Chimie, Universite de Strasbourg, France, for supplying the experimental dipole moment data. (35) N. Bauer, K. Kajans, and 5. Z. Lewin, ”Physical Methods of Organic Chemistry,” Vol. I, Part 2, A. Weissberger, Ed., Interscience Publishers, New York, N. Y., 1960, pp 1254-1270. (30) I. F. Halverstadt and W. D. Kumler, J . Amer. Chem. Soc., 64, 2988 (1942). (37) E. A. Guggenheirn, Trans. Faradav Soc., 46, 714 (1949). (38) J. W. Smith, “Electric Dipole Momenta,” ButterNorth and Co., Ltd., London, 1955.
Dienones Derived from 5-Methoxy[2.2]metacyclophanes~ V. BOEKELHEIDE, C. RAMEY,E. STURM,T. MIYASAKA, AND B. A. HESS,JR. Department of Chemisty, University of Oregon, Eugene, Oregon 9’7403 Received August 28, 1968
Treatment of 5,l3-dimethoxy[2.2]metacyclophane (4) under the conditions of the Rieche reaction gives predominantly the corresponding 8-formyl derivative ( 5 ) . Attempts to convert 5 into a methylene-bridged cis[2.2]metacyclophane were unsuccessful. However, 5 could be converted into the corresponding %methyl derivative (7) and this in turn was oxidized to the Corresponding dienone (8). Also, it has been shown through the preparation of 5-methoxy-8-methyl[2.2]metacyclophane (11) that the Wurts reaction can be used effectively to prepare “mixed dimeric” products. Oxidation of 11 gave the corresponding dienone 12 and treatment of 12 with acetic anhydride and perchloric acid readily effected a dienone-phenol type of rearrangement to give 2-acetoxy-3-methyl-4,5,9,1O-tetrahydropyrene(13).
The synthesis of tran~-l5,16-dimethyldihydropyrene~~~ ( 1) and trans-15,16-diethyldihydr~pyrene~ (2) have
demonstrated the possibility of preparing aromatic molecules having substituents within the cavity of the aromatic ?r cloud of electrons. In a continuation of this study one of the molecules of high interest would be the corresponding trans-15-methyldihydropyrene (3). Having hydrogen as one of the substituents in the cavity of the ?r cloud would open the possibility of removal of the “internal” hydrogen t o form the corresponding anion, cation, or radical. Such species would not only be inherently interesting but might allow the direct introduction of various other substituents. Also, if the internally substituted hydrogen were labile, it might allow its equilibration to the still unknown ~is-15~16-dihydropyrene system. For these reasons we undertook a study of the synthesis of trans-15-methyldihydropyrene (3). (1) This work was supported in part by the Office of Naval Research and in part by the National Science Foundation. (2) V. Boekelheide and J. B. Phillips, PTOC.Noli. Accrd. Sei. U. S.,61. 550 (1964).
V.
Boekelheide and J. B. Phillips, J . Amer. Chem. Soc., 89, 1695 (1967). (4) V. Boekelheide and T. Miyasaka, (bid., 89, 1709 (1967).
(3)
@ \
1, R = CH, 2, R = C,Hj
3
The synthetic approach envisioned was modeled closely to the route proven successful for 1 and 2. This required in the first instance the synthesis of 5,13dimethoxy-8-methyl[2.2]metacyclophane (7). Since the Wurtz reaction, which is the standard procedure for preparing [2.2]metacyclophanes, is normally utilized only for the synthesis of symmetrical molecules, a modification of this approach was necessary. One possibility was t o subject the readily available 5,13dimethoxy[2.2]metacyclophane5 (4) to electrophilic substitution in the hope that substitution would occur a t the 8 position in reasonable yield. Actually, examination of models or a projection drawing such as (5) V. Boekelheide and R. W. Griffin, Jr., J . Orp. Chem., 34, 1960 (1909).
